PROPERTIES OF THE NERVE CELL. 131 



that the axis cylinder can be formed only as an outgrowth from a nerve cell. 

 Some histologists Apathy, Bethe, Nissl have also attacked the most 

 fundamental feature of the neuron doctrine the view, namely, that each 

 neuron represents an independent anatomical element. These authors 

 contend that the neurofibrils of the axis cylinder pass through the nerve cells 

 and enter by way of a network into direct connection with the neurofibrils 

 of other neurons (see Fig. 64). The neurofibrils form a continuum through 

 which nerve impulses pass without a break from neuron to neuron. Ac- 

 cording to this conception, the ganglion cells play no direct part in the con- 

 duction of the impulse from one part of the nervous system to another; 

 the neurofibrils alone, and the intra9ellular and pericellular networks with 

 which they connect, form the conducting paths that are everywhere in con- 

 tinuity. In the explanation given below of the activities of the nervous 

 system, the author, following the usual custom, makes use of the neuron 

 doctrine. 



The Varieties of Neurons. The neurons differ greatly in 

 size, shape, and internal structure, and it is impossible to classify 

 them with entire success from either a physiological or an anatomical 

 standpoint. Neglecting the unusual forms whose occurrence is 

 limited and whose structure is perhaps incompletely known, there 

 are three distinct types whose form and structure throw some 

 light on their functional significance: 



I. The bipolar cells. This cell is found in the dorsal root gan- 

 glia of the spinal nerves and in the ganglia attached to the sensory 

 fibers of the cranial nerves, the ganglion semilunare (Gasserian) 

 for the fifth cranial, the g. geniculi for the seventh, the g. vestibu- 

 lare and g. spirale for the eighth, the g. superius and g. petrosum 

 for the ninth, the g. jugulare and g. nodosum for the tenth. 



The typical cell of this group is found in the dorsal root ganglia. 

 In the adult the two processes arise as one, so that the cell seems to 

 be unipolar, but at some distance from the cell this process divides 

 in T, one branch passing into the spinal cord via the posterior 

 root, the other entering the spinal nerve as a sensory nerve fiber 

 to be distributed to some sensory surface. Both processes become 

 medullated and form typical nerve fibers. That these apparently 

 unipolar cells are really bipolar is shown not only by this division 

 into two distinct fibers, but .also by a study of their development 

 in the embryo. In early embryonic life the two processes arise 

 from different poles of the cell, and later become fused into an ap- 

 parently simple process (Fig. 60). The striking characteristics of 

 this cell, therefore, are that it gives rise to two nerve fibers, and that 

 it possesses no dendritic processes. On the physiological side these 

 cells might be designated as sensory cells, since they appear to be 

 associated always with sensory nerve fibers. 



The nerve cells found in the sensory ganglia exhibit, as a matter of fact, 

 a number of different types, some of which possess short dendritic processes. 

 These histological variations cannot as yet be given a physiological signifi- 

 cance, but their occurrence certainly seems to indicate a possibility that 

 the sensory ganglia may have a much more varied physiological activity 



